rf and the sound of scilence

[andrew s]

There has been some discussion on rf and its impact and mitigation so I did some experiments.  My layout is not optimal.  The bt hub is 2m from my rig and an ethernet over mains is in the same mains socket as the rig and network hub, bridge and NAS.

In addition the i7 NUC is at the bottom of the rack about 0.7m from the DAC.

Kit is as in my WAM details. 

So the test was to play from the NAS a silent track followed by a 1k tone produced in audacity. This was done 1) with the minimum running to play the tracks and 2) as 1 plus streaming Netflix over the mains ethernet,  my mobile phone on a call sitting on the DAC and streaming YouTube to my tablet resting on the power amp.

Results.

1) absolute scilence even at max volume (ear up against base then coincident driver).

Pure 1k tone from listening position no hint of harmonics or other distortion at 70dBA.

2) exactly as 1)

Some years ago I decided not to train my listening but tracks I know well sounded (as far as I can tell) the same in conditions 1 and 2. Obviously,  this was not done blind!

Conclusion 

My rig is immune to rf in my environment with my ears as test instrument.

Regards Andrew 

 

[andrew s]

I should have added no special cables, PC world or home made Cat6 ethernet, £10 -£20 interconnects, standard mains cables and Maplins best speaker cables.

Regards Andrew 

 

[tuga]

There has been some discussion on rf and its impact and mitigation so I did some experiments.  My layout is not optimal.  The bt hub is 2m from my rig and an ethernet over mains is in the same mains socket as the rig and network hub, bridge and NAS.

In addition the i7 NUC is at the bottom of the rack about 0.7m from the DAC.

Kit is as in my WAM details. 

So the test was to play from the NAS a silent track followed by a 1k tone produced in audacity. This was done 1) with the minimum running to play the tracks and 2) as 1 plus streaming Netflix over the mains ethernet,  my mobile phone on a call sitting on the DAC and streaming YouTube to my tablet resting on the power amp.

Results.

1) absolute scilence even at max volume (ear up against base then coincident driver).

Pure 1k tone from listening position no hint of harmonics or other distortion at 70dBA.

2) exactly as 1)

Some years ago I decided not to train my listening but tracks I know well sounded (as far as I can tell) the same in conditions 1 and 2. Obviously,  this was not done blind!

Conclusion 

My rig is immune to rf in my environment with my ears as test instrument.

Regards Andrew 

Just out of curiosity, can you try two tones, perhaps 19kHz and 20kHz at -6dBFS?

 

[Fourlegs]

There has been some discussion on rf and its impact and mitigation so I did some experiments.  My layout is not optimal.  The bt hub is 2m from my rig and an ethernet over mains is in the same mains socket as the rig and network hub, bridge and NAS.

In addition the i7 NUC is at the bottom of the rack about 0.7m from the DAC.

Kit is as in my WAM details. 

So the test was to play from the NAS a silent track followed by a 1k tone produced in audacity. This was done 1) with the minimum running to play the tracks and 2) as 1 plus streaming Netflix over the mains ethernet,  my mobile phone on a call sitting on the DAC and streaming YouTube to my tablet resting on the power amp.

Results.

1) absolute scilence even at max volume (ear up against base then coincident driver).

Pure 1k tone from listening position no hint of harmonics or other distortion at 70dBA.

2) exactly as 1)

Some years ago I decided not to train my listening but tracks I know well sounded (as far as I can tell) the same in conditions 1 and 2. Obviously,  this was not done blind!

Conclusion 

My rig is immune to rf in my environment with my ears as test instrument.

Regards Andrew 

Unfortunately your experiments are rather (completely) flawed and you are wrong to claim that your rig is immune to RF in your environment and with your ears.

All the talk of RF noise is because RF noise can and often does get into the analogue stages of the DAC causing intermodulation distortion. This is easily heard as a hardening or emphasis to the treble but can also cause bass and lower mids to lose detail. 

Your first test with silence will not and cannot pick up any of the RF noise because its frequency is far outside the audible range. The intermodulation distortion caused by the RF requires music to be playing for the distortion to take place. No music being played means there is no distortion.

With your second test with a 1k tone I doubt whether any of the RF noise distortion artefacts would be audible because they are probably swamped but the dominant pure tone.

Your best test as to whether your system has any issues is to play a favourite track and then take some RF mitigation measures and listen for whether you hear any difference in the treble. Mitigation might be the use of ferrited cables between your source and your dac or it might be moving your system further away from RF sources such as the router or it might be something else.

 

[andrew s]

Unfortunately your experiments are rather (completely) flawed and you are wrong to claim that your rig is immune to RF in your environment and with your ears.

All the talk of RF noise is because RF noise can and often does get into the analogue stages of the DAC causing intermodulation distortion. This is easily heard as a hardening or emphasis to the treble but can also cause bass and lower mids to lose detail. 

Your first test with silence will not and cannot pick up any of the RF noise because its frequency is far outside the audible range. The intermodulation distortion caused by the RF requires music to be playing for the distortion to take place. No music being played means there is no distortion.

With your second test with a 1k tone I doubt whether any of the RF noise distortion artefacts would be audible because they are probably swamped but the dominant pure tone.

Your best test as to whether your system has any issues is to play a favourite track and then take some RF mitigation measures and listen for whether you hear any difference in the treble. Mitigation might be the use of ferrited cables between your source and your dac or it might be moving your system further away from RF sources such as the router or it might be something else.

Sorry to disagree.

As I said I did play my favourite tracks and did not hear any difference.  One can't mitigate what does not exist. The "beauty" of subjective test is they are irrefutable by definition.  If I had a spectrum analyser I would do it objectively but unfortunately I don't.  

rf could cause problems other than the IM of music you claim. Various mechanisms could demodulate it or fold it I to the audible. If I put my phone near my computer I can clearly hear it create noise.

@tuga I will try your test.

Regards Andrew 

 

[Shadders]

Sorry to disagree.

As I said I did play my favourite tracks and did not hear any difference.  One can't mitigate what does not exist. The "beauty" of subjective test is they are irrefutable by definition.  If I had a spectrum analyser I would do it objectively but unfortunately I don't.  

rf could cause problems other than the IM of music you claim. Various mechanisms could demodulate it or fold it I to the audible. If I put my phone near my computer I can clearly hear it create noise.

@tuga I will try your test.

Regards Andrew 

Hi,

Here is a good link :

https://trinity-ed.de/problems-with-intermodulation-distortions/

The intermodulation between RF and audio is in general, always out of band. The simple equations in the link above if you use 30MHz and 19kHz will show that you will need to take into account the 1,579th order before it is in band, and at this order, the magnitude is near non-existent.

What can happen is demodulation of RF into the audio band, and this happens with or without any music. If you hear nothing, then it is not happening in your equipment at a level you can hear. It is probably not happening at all.

Regards,

Shadders.

 

[Fourlegs]

Sorry to disagree.

As I said I did play my favourite tracks and did not hear any difference.  One can't mitigate what does not exist. The "beauty" of subjective test is they are irrefutable by definition.  If I had a spectrum analyser I would do it objectively but unfortunately I don't.  

rf could cause problems other than the IM of music you claim. Various mechanisms could demodulate it or fold it I to the audible. If I put my phone near my computer I can clearly hear it create noise.

@tuga I will try your test.

Regards Andrew 

If you are happy and content then that is really all that matters. xD

 

[andrew s]

So I have tried the 20k and 19k tones. Both with and without extra rf. 

In both cases I could hear the two tones plus an extremely faint hum at 50 ish Hz. No sign of any 1k tone.

My room has modes close to 50 Hz but it was coming from the speakers albeit a a very low level. 

Regards Andrew 

 

[Radioham]

Put your mobile phone close to various pieces of the hifi and you will soon see how immune the system is. Try send or receiving a text to show that its not idle.

Alan

 

[MF 1000]

Put your mobile phone close to various pieces of the hifi and you will soon see how immune the system is. Try send or receiving a text to show that its not idle.

Alan

There used to be a taxi that went past our house (set 30 ft back from the road) and if he transmitted on his radio my DHT preamp would pick up the radio signal ! 

 

[andrew s]

Put your mobile phone close to various pieces of the hifi and you will soon see how immune the system is. Try send or receiving a text to show that its not idle.

Alan

I did. It was on a call and placed on top of the DAC and it was not picked up.

There used to be a taxi that went past our house (set 30 ft back from the road) and if he transmitted on his radio my DHT preamp would pick up the radio signal ! 

Happened to me years ago and I nearly had a heart attack.

Regards Andrew 

 

[Griff500]

Doesn't the word 'noise' cause confusion and make people think they should be able to hear it, rather than hearing the benefits of removing it?

 

[Fourlegs]

Doesn't the word 'noise' cause confusion and make people think they should be able to hear it, rather than hearing the benefits of removing it?

Yes. Often it is in the 2 to 5GHz region. 

 

[andrew s]

Doesn't the word 'noise' cause confusion and make people think they should be able to hear it, rather than hearing the benefits of removing it?

But surely, you must be able to hear some change in the audible. Lower noise floor for example.

What examples would you give of "hearing the benefit of removing it" and how would that differ from hearing it before it was removed?

To me I one is the inverse if the other or am I missing something? 

Regards Andrew 

 

[Fourlegs]

But surely, you must be able to hear some change in the audible. Lower noise floor for example.

What examples would you give of "hearing the benefit of removing it" and how would that differ from hearing it before it was removed?

To me I one is the inverse if the other or am I missing something? 

Regards Andrew 

The RF noise is inaudible unless there is music playing. When the music is playing you might just think, oh, that has a wide sound stage or that has a very clear presentation or  that sounds very 3D or possibly (if you know what to listen for) you might think the top end sounds a bit hard or over emphasised and that the music was slightly fatiguing.

Remove the RF noise and generally everything sounds a bit smoother and less fatiguing. Instruments and notes actually have more real detail because they are not masked by the RF noise IMD artefacts. But unless you are familiar in what to listen out for it takes the removal of the RF noise before one realises that the problem was there.

That is what Griff means about hearing the benefits from removing it.

 

[Griff500]

But surely, you must be able to hear some change in the audible. Lower noise floor for example.

What examples would you give of "hearing the benefit of removing it" and how would that differ from hearing it before it was removed?

To me I one is the inverse if the other or am I missing something? 

Regards Andrew 

I'm no expert in the technical side of these things, or in describing sound quality, but I'll have a go (and might get it wrong).  xD

I think using the word 'noise' makes it seem like something you should be able to hear. 'Interference' is the word and when it is removed, such as when using WAVE Storm cables to address specific frequencies, I seem to hear more detail and it feels like there is more space between instruments. I also find that I can listen longer at higher volume levels without feeling the need to turn the volume down.

 

[andrew s]

The RF noise is inaudible unless there is music playing. When the music is playing you might just think, oh, that has a wide sound stage or that has a very clear presentation or  that sounds very 3D or possibly (if you know what to listen for) you might think the top end sounds a bit hard or over emphasised and that the music was slightly fatiguing.

Remove the RF noise and generally everything sounds a bit smoother and less fatiguing. Instruments and notes actually have more real detail because they are not masked by the RF noise IMD artefacts. But unless you are familiar in what to listen out for it takes the removal of the RF noise before one realises that the problem was there.

That is what Griff means about hearing the benefits from removing it.

Not questioning what you hear but how do you know it is down to RF noise IMD artefacts? It would seem to me very difficult to measure these in a music stream spectrogram or other measurement. 

Regards Andrew 

 

[Fourlegs]

Not questioning what you hear but how do you know it is down to RF noise IMD artefacts? It would seem to me very difficult to measure these in a music stream spectrogram or other measurement. 

Regards Andrew 

I admit I rely to a great extent on the expertise and experience of Rob Watts who has investigated this in some detail with the benefit of his technical knowledge and sensitive measuring devices which he has available to use. He has written many times on this subject. He is not the only designer who is conversant with this and I would say it is now fairly standard and well understood stuff. (That still doesn't stop some lesser informed people think they know better xD .)

 

[Griff500]

I must admit that I did think the improvements in sound quality from reducing RFI were well established and/or widely accepted based on what I have read "on the line".